The influence of hardware and software considerations on the design of synchronous parallel algorithms

30 October 1983

journal article
research article
Published by Wiley in Software: Practice and Experience

Vol. 13 (10) , 961-974
https://doi.org/10.1002/spe.4380131007

Abstract

Matrix multiplication is used as an example to illustrate a method of transforming the specification of a problem into an algorithm suitable for execution on synchronous machines. The transformations are influenced both by the architectures of the target machines and their available high level languages. Three different synchronous machines, a conceptual MCC (mesh connected computer), the Cray‐1 and the ICL DAP are considered as target hardware.

Keywords

This publication has 6 references indexed in Scilit:

Non-numerical methods on parallel computers
Computer Physics Communications, 1982
Parallel Matrix and Graph Algorithms
SIAM Journal on Computing, 1981
An optimal routing algorithm for mesh-connected Parallel computers
Journal of the ACM, 1980
A Language for Array and Vector Processors
ACM Transactions on Programming Languages and Systems, 1979
The CRAY-1 computer system
Communications of the ACM, 1978
The programming language pascal
Acta Informatica, 1971